Flagship Brand Stores within Virtual Worlds: The

advertisement
Recherche et Applications en Marketing, vol. 24, n° 3/2009
Flagship Brand Stores within Virtual Worlds:
The Impact of Virtual Store Exposure on Real-Life Attitude toward
the Brand and Purchase Intent
Michael Haenlein
Professor of Marketing
ESCP Europe
Andreas M. Kaplan
Professor of Marketing
ESCP Europe
ABSTRACT
Virtual hyperrealities, also referred to as virtual social worlds, have experienced increasing managerial interest in recent years.
Although they have also received some attention in the academic literature, the extent to which corporate presences within such
environments can influence attitude toward the brand and purchase intent in real life remains unclear. Based on a survey
conducted among 580 Second Life residents, we show that exposure to flagship brand stores within virtual worlds positively
influences attitude toward the associated brand and real life purchase intent. We furthermore show that a user’s purchase
experience (shopping frequency, purchase frequency, spending per purchase) and the gratification derived from the use of
their purchases have a significant moderating effect on these relationships. Our results are of managerial and theoretical
importance as they provide empirical evidence for spill-over effects between virtual worlds and real life and help to develop
recommendations on optimal store design within virtual social worlds.
Key words: Avatar, attitude toward advertising, attitude toward the brand, flagship store, hyperreality, Second Life, social
media, virtual social world, virtual world.
The authors contributed equally to this work and are listed in alphabetical order. The authors thank the managers from the market research firm,
Repères, especially François Abiven and Emilie Labidoire, for their help during data collection.
The authors may be contacted at the following e-mail addresses: haenlein@escpeurope.eu; mail@andreaskaplan.eu
58
Michael Haenlein, Andreas M. Kaplan
INTRODUCTION
Since its rise in the late 1960s, postmodernism
has been forecast to have a strong impact on the
Marketing discipline. As discussed, for example, by
Firat and Venkatesh (1993) and illustrated by Vargo
and Lusch (2004), the basic conditions of this philosophical stream, such as fragmentation or the reversal
of production and consumption, are closely related to
the core domains of Marketing. Yet, surprisingly,
only a few papers with a postmodern focus have been
published in leading Marketing journals over the past
40 years (for a notable exception, see two special
issues of the IJRM in 1993 and 1994, Firat, Sherry
and Venkatesh, 1994; Venkatesh, Sherry and Firat,
1993). Nevertheless, postmodern ideas have increasingly spread into the business world. Examples of
this evolution include the rising use of hyperrealities,
i.e., artificially created settings that appear to be real
for the individuals involved in them, in the context of
tourist attractions (Grayson and Martinec, 2004), reality television shows (Rose and Wood, 2005), and
retail settings (Edvardsson, Enquist and Johnston,
2005). As discussed, for example, by Edvardsson,
Enquist and Johnston (2005), IKEA, one of the largest and most successful retail furniture companies
in the world, uses hyperreal experience rooms as a
strategic tool to create favorable and memorable prepurchase service experiences.
Recently, hyperrealities have started to become
even more important due to the creation of their virtual
equivalents in the online world. One of the more prominent examples of these virtual hyperrealities,
which are also referred to under the term virtual
social worlds, is probably the online application
Second Life (SL). Second Life is a three-dimensional
virtual social world which users (who are called
“residents”) can enter through a downloadable client
program. Second Life allows its residents to interact
with one another in the form of personalized avatars,
explore their environment and sell or buy virtual products and services. While some of these products and
services are available free of charge, the higher quality
and branded versions are usually sold for virtual
money (Linden Dollars, L$). Such money can,
among others, be obtained by exchanging real-life
currencies via the Second Life Exchange at a floating
exchange rate that is approximately stable at 250 L$ to
the US$ or, alternatively, be earned in-world through
jobs or interest payment (Kaplan, 2009). The high
popularity of Second Life has recently motivated
several real-life companies to start activities within
Second Life, including Dell, Toyota, and American
Apparel. These firms decided to set up virtual flagship brand stores within Second Life (called
“islands”), which are subsequently used as hubs for
brand communities and the distribution of virtual
products and services.
Second Life in particular and virtual worlds in
general have received some interest in the academic
literature in recent years (see, for example, Barnes
and Mattsson, 2008; Kaplan and Haenlein, 2009a;
Kaplan and Haenlein, 2009b; Kaplan and Haenlein,
2009c; Lin, 2008; Wood, Solomon and Allan, 2008;
Yee et al., 2007), and since 2008 there has even been a
journal specifically dedicated to this subject (Spence,
2008). Although the hype around this topic has now
come to an end, the market research company
Gartner still estimates that by 2012, 70% of all organizations will have established their own private virtual
world. Our analysis focuses on one question that has
received less attention so far, namely the extent to
which corporate presences within Second Life can
influence brand attitudes and purchase intent in real
life. Based on a survey among 580 Second Life residents, we show that exposure to Second Life flagship
stores impacts real-life brand attitudes and purchase
intent. Specifically, we observe a positive and significant relationship between a user’s attitude toward a
Second Life flagship store and his/her attitude toward
the real-life brand and his/her attitude toward the
real-life brand and real life purchase intent. In addition,
our results indicate that a user’s purchase experience
and use gratification significantly moderate the relationship between store attitude antecedents, attitude
toward a Second Life flagship store and attitude
toward the real-life brand. This shows that different
types of residents are likely to react differently to
exposure to virtual flagship stores – a characteristic
that needs to be accounted for when designing flagship
stores within this environment.
Flagship Brand Stores within Virtual Worlds
CONCEPTUAL FRAMEWORK
Our conceptual framework builds on the assumption that exposure to a flagship store within Second
Life (or any virtual social world for that matter)
impacts real-life brand attitude and purchase intent.
We therefore hypothesize that Second Life flagship
stores fulfill functions similar to other forms of
(online) advertising, besides their primary purpose of
serving as hubs for the distribution of virtual products and services. There are at least three different
lines of thinking that can be used to justify this basic
premise: First, it has been discussed in the retailing
literature that companies install traditional (physical)
flagship brand stores to achieve benefits that go
beyond their basic role as points of sale. As highlighted, for example, by Kozinets et al. (2002,
p. 17), flagship brand stores, which are defined as
outlets that are owned by the manufacturer and carry a
single (usually established) brand of product only,
are operated at least partly “with the intention of
building or reinforcing the image of the brand rather
than operating to sell products at a profit.” Using the
example of the Milan-based furniture manufacturer
B&B Italia, Doyle et al. (2008) demonstrate the
value of flagship stores as a brand management
device in the context of the Italian luxury furniture
sector. In a similar spirit, Brakus, Schmitt and
Zarantonello (2009) provide empirical evidence that
the shopping and service experience which is delivered
in such stores represents an essential part of the overall
brand experience. Given that a central role of traditional advertising is equally to create favorable attitudes toward the brand (e.g., Shimp, 1981), this
implies that traditional flagship brand stores fulfill
the dual function of serving as an advertising
medium and as a sales outlet. We assume that a similar
thinking can be applied when analyzing flagship
stores within virtual worlds, such as Second Life.
Second, it has been shown in the advertising field
that online advertising can be an effective tool to
influence brand attitudes and purchase intent both
online and offline. Manchanda et al. (2006) analyze,
for example, the impact of banner advertising on
Internet purchasing and show a positive impact on
repeat purchase probabilities. In a recent literature
review, Ha (2008) summarizes more than 80 articles in
59
leading advertising journals that have addressed the
area of online advertising (e.g., banner ads, pop-up
ads, rich media ads) since 1996. Consistently, these
studies show that online advertising can lead to positive brand evaluations and tends to be more efficient in
achieving this objective than traditional offline print
advertising. It is therefore reasonable to assume that
advertising on the Internet (be it in the form of traditional online advertising or of flagship stores within
virtual worlds) can spill over to the real world and
influence offline consumer behavior.
Finally, research in the area of consumer behavior
has provided consistent support for the fact that
consumers consider their activities within the online
space as an integral part of their life and that there is a
transfer of skills learned offline to the online world
and vice versa. Yee and Bailenson (2007), for
example, show that behavioral patterns familiar from
the real world are equally applied in virtual settings – a
phenomenon which they refer to as the Proteus
effect. In a series of experiments, these authors provide
evidence that subjects who are represented by more
attractive avatars tend to be more intimate with
confederates in a self-disclosure and interpersonal
distance task and that subjects who are represented
by taller avatars behave more confidently in a negotiation task. In a similar spirit, Schlosser (2003; 2006)
shows that experiencing products in virtual environments leads to more favorable attitudes and higher
purchase intentions due to higher object interactivity.
Among others, these studies can be seen as an indication that consumers do not separate their online from
their offline identity but instead use media such as
Second Life or personal web pages to create an alternative presentation of themselves within the virtual
space (Schau and Gilly, 2003). Given that both environments are therefore closely related in the mind of
the consumer, it is likely that brand exposures and
experiences made in virtual worlds will have an
impact on brand evaluations and attitudes in real life.
60
Michael Haenlein, Andreas M. Kaplan
RESEARCH MODEL AND HYPOTHESIS
DEVELOPMENT
shopkeeper, attitude toward purchasing on Second
Life and mood. This leads to the following set of
hypotheses:
H1: A user’s attitude toward a Second Life flagship store is positively influenced by (a) store credibility, (b) attitude toward the shopkeeper, (c)
attitude toward purchasing on Second Life and
(d) mood.
As highlighted above, we assume that exposure to
a Second Life flagship store can influence real-life
brand attitudes and purchase intent in a similar way as
traditional advertising. To investigate this hypothesis,
we apply a modification of the well-established
attitude toward the ad – attitude toward the brand
(AAd – ABrand) model to our research setting (See
Figure 1). The relationship between AAd and ABrand
has been investigated regularly in advertising literature, starting with the seminal papers of Shimp
(1981) and Mitchell and Olson (1981). Consistently
these studies support a causal relationship between
AAd – ABrand and ABrand – purchase intent. In an analogous way, we assume that a Second Life resident’s
attitude toward a Second Life flagship store (AStore)
influences his/her attitude toward the real-life brand
which subsequently drives real-life purchase intent.
In this context we define a Second Life resident as a
user who owns at least one Second Life account and
uses this account on a regular basis. Regarding the
antecedents of AAd, MacKenzie and Lutz (1989) discuss that ad credibility, attitude toward the advertiser,
attitude toward advertising and mood jointly
influence attitude toward the ad. The influence of ad
credibility on AAd is consistent with literature in the
area of source credibility theory stating that information from credible sources tends to be more effective
than from incredible ones (e.g., Hovland and Weiss,
1951). The influence of attitude toward advertising
and toward the advertiser on AAd is consistent with a
process of generalization in which consumers’ general
affective reactions influence attitudes toward specific
advertising messages. The impact of mood on AAd is
consistent with research showing that transient feeling states that are subjectively perceived by individuals (i.e., mood) have a substantial influence on
consumer behavior (See Gardner, 1985b, for a
review). In analogy, we assume that AStore is
influenced by store credibility,1 attitude toward the
Within this general framework, we assume that
the relative importance of different variables is not
the same for different types of Second Life residents.
Specifically, we hypothesize a moderating role of a
Second Life resident’s usage intensity, purchase
experience, and use gratification on the aforementioned relationships. With respect to usage intensity and
purchase experience, it has been shown that repeated
exposure to an activity leads to different types of
behavior due to mechanisms such as conditioning
(e.g., McSweeney and Bierley, 1984; Stuart, Shimp
and Engle, 1987) or the use of scripts as a basis for
decision making (e.g., Wofford and Goodwin, 1990).
Exemplary studies that support such thinking in an
online context include Gefen, Karahanna and Straub
(2003) for e-Commerce, Kang et al. (2006) for the
use of e-coupons, Qiu and Papatla (2008) for the
acquisition of free online content, and Lemmens and
Bushman (2006) for the emotional consequences of
violent video game consumption. Within our
research framework, we expect for example that store
credibility as well as attitude toward the shopkeeper or
purchasing on Second Life can only build up over
time and, therefore, are likely to play a different role in
influencing store attitudes for low versus high levels of
usage intensity/purchase experience. More generally,
we postulate that the relative importance of different
store attitude antecedents as well as the strength of
1. According to Erdem and Swait (2004), credibility can be defined
as the believability of an entity’s intentions at a particular time.
Credibility consists of two main components which express the
ability (i.e., expertise) and willingness (i.e., trustworthiness) to
deliver what has been promised. As shown by Erdem and Swait
(2004), brand credibility increases the probability of the inclusion of
a brand into the consumer’s consideration set as well as brand
choice conditional on consideration. Our construct of “store credibility” extends this idea of brand credibility (or source credibility in
general) to flagship stores within virtual worlds.
H2: A user’s attitude toward a Second Life flagship store positively influences his/her attitude
toward the real-life brand.
H3: A user’s attitude toward the real-life brand
positively influences his/her real-life purchase
intent.
Virtual world resident
usage intensity
H4
Attitude toward
the shopkeeper
H1c
Attitude toward
the Second Life
flagship store
H1b
Attitude toward
purchasing on
Second Life
H1d
H2
H1a
H3
Figure 1. – Research model and hypotheses
Virtual world resident
purchase experience
H5
Mood
Attitude toward
the Real Life
brand
Store credibility
Virtual world resident
use gratification
H6
Real Life
purchase intent
Flagship Brand Stores within Virtual Worlds
61
62
Michael Haenlein, Andreas M. Kaplan
the relationship between store and brand attitude
change with increasing levels of usage intensity (e.g.,
usage duration, weekly usage) and purchase experience (i.e., shopping frequency, purchase frequency,
and spending per purchase). This leads to the following two hypotheses:
H4: A user’s usage intensity of Second Life will
have a moderating effect on the impact of (a)
store attitude antecedents on the attitude toward
a Second Life flagship store and (b) attitude
toward a Second Life flagship store on the attitude toward the real-life brand.
H5: A user’s purchase experience within Second
Life will have a moderating effect on the impact of
(a) store attitude antecedents on the attitude
toward a Second Life flagship store and (b) attitude
toward a Second Life flagship store on the attitude toward the real life brand.
Regarding the moderating impact of use gratification, it has been discussed regularly in mass media
research that consumers use media to fulfill certain
needs, i.e., to satisfy certain kinds of gratification
(See Katz, Blumler and Gurevitch, 1973-74, for an
early summary of uses and gratification research). In
one of the first studies in this area, McQuail, Blumler
and Brown (1972) applied a uses and gratification
perspective to the analysis of television quiz shows
and showed that self-rating, excitement, and educational appeals can be used to explain the consumption of such programs. In the last few decades, use
gratification has regularly been applied to investigate
factors driving the consumption of new media.
Exemplary recent studies include Stafford, Stafford
and Schkade (2004), who investigate gratification of
Internet usage, and Chung and Kim (2008) who
applied this approach to the analysis of blog use by
cancer patients. In the area of virtual worlds, Kaplan
and Haenlein (2009a; 2009b) have identified four
key motivations for Second Life usage: the search for
diversion, the desire to build personal relationships,
the need to learn, and the wish to earn money. In our
analysis, we assume that the gratification that Second
Life users intend to achieve with their usage will
have a moderating effect on the relationships expressed
in H1 and H2. Specifically, we postulate that the
motivations for Second Life usage influence the relative importance of different store attitude antecedents
as well as the strength of the relationship between
store and brand attitude. This assumption is consistent with the work of Liang, Lai and Ku (2006-2007)
who have shown that Internet use gratification moderates user satisfaction with personalized online services, as well as with Ko, Cho and Roberts (2005)
who provide evidence that use gratification
influences Internet users’ attitudes toward interactive
advertising sites. Combined, this leads to the following hypothesis:
H6: A user’s use gratification of Second Life will
have a moderating effect on the impact of (a)
store attitude antecedents on the attitude toward
a Second Life flagship store and (b) attitude
toward a Second Life flagship store on the attitude toward the real life brand.
RESEARCH METHODOLOGY
We decided to test the aforementioned hypotheses using a quantitative (vs. qualitative) study as
this approach allows for the use of formal statistical
procedures to verify the accuracy of our predictions
and additionally provides us with an estimate of the
strength of the effects hypothesized within our
research model. Within the field of quantitative
methods, we opted for a survey-based approach,
which ensures a higher degree of external validity
than other alternatives, such as, for example, laboratory
experiments.
Measurement scales and Data collection procedure
Within our conceptual framework, we operationalized store credibility using an adaptation of the
brand credibility scale developed by Erdem and
Swait (2004), including dimensions of expertise and
trustworthiness. For the attitude toward purchasing
on Second Life, we relied on Iyer and Eastman
(2006) and adapted their measure of attitude toward
e-Commerce. Attitude toward the shopkeeper and
attitude toward the Second Life flagship store were
operationalized using an adaptation of items used by
MacKenzie and Lutz (1989). For mood, we used a
Flagship Brand Stores within Virtual Worlds
four-item scale proposed by Swinyard (1993).
Attitude toward the real life brand relied on a threeitem scale proposed by Gardner (1985a) and real life
purchase intent was measured as suggested by Baker
and Churchill (1977). Use gratification was measured with scales derived from McQuail et al. (1972),
building on the work of Kaplan and Haenlein (2009a;
2009b). Following the recommendations of Cox
(1980), all items were measured on 7-point Likert
scales with the default response cues being strongly
disagree (–3) and strongly agree (+3). All scales (See
Appendix for details) were pre-tested using a sample
of 22 students, which resulted in mild changes in
wording, and proved to have good reliability in our
final sample.
Since the nature of Second Life makes it difficult
to obtain a sufficiently large and representative
sample of potential survey participants, we collaborated with a market research firm (Repères) that
maintains a representative panel of 10,000 Second
Life residents. To build the panel, Repères started to
install opt-in terminals at different locations in
Second Life (e.g., residential areas, shopping malls,
night clubs) and set up an affiliate program with
well-established Second Life residents in October
2006. Using official user statistics provided by
Linden Research Inc., the operator of Second Life,
Repères regularly verifies that their panel is representative of the Second Life population in terms of key
user demographics (e.g., gender, age, country of origin). For our study, we contacted a subset of this
panel with the task of visiting one of five pre-selected
Second Life flagship stores. This store list was obtained by matching the Interbrand Best Global Brands
Ranking with the directory of corporate presences in
Second Life, paying special attention to including
stores from different sectors and contained the Dell
Factory on Dell Island, the Mercedes Benz Island,
the Nike Main Store, the Philips Design Island, and the
Sony BMG Island.2 Panelists were subsequently
2. Specifically, we relied on the 2007 Best Global Brands Ranking
which includes the 100 most valuable brands worldwide. First, we
checked for the Top 50 brands whether an official corporate presence existed within Second Life. This resulted in a total of 13
matches. Second, we identified the industry sectors these 13
brands belonged to (Computer services, hardware and software: 6,
Automotive: 3, Consumer electronics, Sporting goods, Diversified
and Beverages: 1). Third, we selected one brand with a Second
Life presence within each of these sectors (Computer services,
hardware and software: Dell; Automotive: Mercedes Benz;
Consumer electronics: Sony BMG; Sporting goods: Nike;
Diversified: Philips).
63
asked to visit one of these five stores and to send
screenshots from their visit back to the market
research company to ensure that all respondents were
exposed to the Second Life flagship store prior to
survey participation. Subsequently, respondents were
contacted a second time with a link to a web page
containing our online survey. Respondents were first
asked to describe their impression of the Second Life
flagship store in one or more sentences (to increase
salience of the stimulus) and then to answer all subsequent questions using this Second Life flagship as a
reference, before they received a compensation to
remunerate them for their effort.
Sample description
Data collection resulted in a total of 717 responses, of which we deleted 65 records due to multiple
submissions of the same respondents and 42 records
due to drop-out before survey completion. Generally,
various studies have shown that data collected
through the Internet does not differ substantially
from other data collection approaches (e.g.,
Birnbaum, 2004; Schillewaert and Meulemeester,
2005). Although web-based studies may suffer from
higher noise due to technical variations, this bias is
usually compensated by the larger sample sizes that
can be achieved through this medium. To minimize
any potential distortions in this respect, we deleted
30 records (5%) that showed particularly high or low
survey response times. This resulted in a final sample
of 580 respondents (80.9%).
Table 1 provides some basic descriptive information regarding our final sample. 28% of our respondents visited the Mercedes Benz Island, 26% the
Sony BMG Island, 24% the Dell Island, 11% the
Nike main store, and 10% the Philips Design Island.
Survey participants are approximately equally split
between men (56%) and women (44%). The majority
(29%) is between 25 and 34 years old, with the
remainder being approximately equally distributed in
the 18-24 years (23%), 35-44 years (26%), and 45+
years (21%) age brackets. With respect to their country
of origin, 26% stem from the USA, 17% from
France, 8% from the UK and Brazil, 5% from
Canada, 4% from the Netherlands and 32% from
other countries. Comparing these characteristics with
general Second Life user statistics based on total
64
Michael Haenlein, Andreas M. Kaplan
Table 1. – Sample description
Sample
Absolute
in %
Gender
Male
Female
Age
13-17 years
18-24 years
25-34 years
35-44 years
45 years and older
Unknown
Country of origin
USA
France
UK
Brazil
Canada
Netherlands
Other
Ethnicity
Asian
Black/African American
Hispanic/ Latino
White
Mixed/ Multi-racial
Other
Occupation
Student/in education
Worker
Employee
Civil servant
Self-employed
Homemaker
Pensioner
Other
Island visited
Mercedes Benz Island
Sony BMG Island
Dell Island
Nike Main Store
Philips Design Island
Second Life population3
% of total hours
% of avatar count
322
258
55.5%
44.5%
58.5%
41.5%
67.2%
32.8%
4
135
167
153
121
0
0.7%
23.3%
28.8%
26.4%
20.9%
0.0%
0.4%
15.4%
34.8%
28.4%
20.6%
0.4%
0.9%
22.8%
35.2%
24.1%
16.5%
0.5%
148
96
49
47
28
24
188
25.5%
16.6%
8.4%
8.1%
4.8%
4.1%
32.4%
38.1%
5.5%
7.2%
3.4%
3.6%
4.0%
38.2%
36.7%
5.0%
7.8%
4.2%
3.3%
3.3%
39.7%
30
13
31
456
31
19
5.2%
2.2%
5.3%
78.6%
5.3%
3.3%
121
78
172
16
101
32
10
50
20.9%
13.4%
29.7%
2.8%
17.4%
5.5%
1.7%
8.6%
162
152
142
65
59
27.9%
26.2%
24.5%
11.2%
10.2%
3. Based on Second Life Virtual Economy Key Metrics (BETA) through April 2008 published by Linden Lab.
Flagship Brand Stores within Virtual Worlds
hours and avatar count at the time of our survey
shows that respondents are similar to the overall
Second Life population average.
In Table 2, we provide some details on Second
Life usage and consumption characteristics within
our sample. Most of our respondents joined Second
Life 6-12 months (41%) or less than 6 months (38%)
ago. The majority of users in our sample spend between 10 and 24 hours (38%) or more than 24 hours
(34%) per week in Second Life, which is equivalent to
1-3 hours per day or more. Combined, this shows
that Second Life represents a substantial part of the
day-to-day activities of the respondents in our
sample. In terms of consumption behavior, the majority of respondents report going shopping once or
several times a week (50%) and making purchases
between once a week and 2-3 times per month
(46%). Product and service purchases are hence an
important part of residents’ Second Life experience.
The total amount of money spent within Second Life
remains, however, reasonably small, with a weighted
average of roughly L$200, or less than US$1 per
week.
Statistical analysis
In order to group respondents into different use
gratification clusters, we first performed a two-step
cluster analysis in SPSS 14.0.4 For this, we used the
twelve use gratification items as categorical indicators and determined clusters based on a log-likelihood distance measure and Schwarz’s Bayesian
Criterion (BIC) as the clustering criterion. This resulted in a three-cluster solution in which Cluster A
accounted for 225 (39%), Cluster B for 249 (43%)
and Cluster C for 106 (18%) respondents. To better
understand the structure of these clusters, we subsequently computed use gratification scores by calculating the (unweighted) average of all items belonging to
the same use gratification and subsequently determined the average score per gratification for each of the
4. The SPSS two-step cluster method is a scalable cluster analysis
algorithm designed to handle very large data sets. It can handle
both continuous and categorical variables and consists of two
steps: first, pre-clustering of all cases into many small sub-clusters; and second, clustering of these sub-clusters into the desired
number of clusters. More details can be obtained from the first
author on request.
65
three clusters. The results of this analysis can be
found in Table 7. As can be seen, Cluster C always
scores higher than Cluster B which again scores
higher than Cluster A for three of the four kinds of
use gratification (diversion, personal relationships,
personal identity; the difference between clusters is
significant: based on F-test, p-value < 0.0005 in all
cases). This can be seen as an indication that the
amount of use gratification obtained through Second
Life usage increases from Cluster A to B and C. We
therefore refer to Cluster C as “heavy” users that rely
on Second Life for various kinds of gratification,
while Clusters B and A are referred to as “medium”
and “light” users, respectively.
We subsequently estimated the structural equation model visualized in Figure 1 using the Mplus
software tool (means, intercepts, and thresholds not
included in the analysis model; information matrix
estimated using second-order derivatives), Version 5
(Muthén and Muthén, 1998-2007).5 To control for
common method bias, we relied on an approach suggested by Podsakoff et al. (2003) and allowed all
items to load on their theoretical constructs as well as
on a latent common method variance factor, uncorrelated to all other model constructs. This resulted in
baseline parameter estimates and standard errors
visualized in Table 3. To test for a potential moderating
impact of usage intensity, purchase experience, and
use gratification, we followed a three-step approach
used by MacKenzie and Spreng (1992). We first estimated group-specific models (covariances, variances,
and residual variances assumed to be equal across
groups) and compared the change in model chisquare between a model in which parameters were
allowed to vary freely across groups and one where
they were constrained to equality. If the constraint
model proved to fit the data significantly worse than
the free model, we considered this as an omnibus
indication of moderation (See panel “omnibus test of
5. Mplus is a software tool for the estimation of generalized latent
variable models, which can be considered as an extension of traditional structural equation models. In the case of our relatively
simple model, the use of Mplus leads to results that are identical to
those that would have been obtained by using other software tools,
such as AMOS or LISREL. Mplus provides, however, substantially more flexibility in other modeling situations. For more
details on generalized latent variable modeling, see Muthén (2002)
and for more details on Mplus, Muthén and Muthén (1998-2007).
66
Michael Haenlein, Andreas M. Kaplan
Table 2. – Sample Second life usage and consumption characteristics
Sample
Absolute
in %
When did you start using Second Life?
Less than 1 week ago
1-4 weeks ago
1-3 months ago
4-6 months ago
6-12 months ago
More than 1 year ago
How many hours do you spend per week in Second Life?
<5 hours
5-9 hours
10-14 hours
15-24 hours
25-44 hours
>45 hours
How often do you go shopping in Second Life?
Never
Less often
2-3 times a month
Once a week
Several times a week
Once a day
Several times a day
How often do you make a purchase in Second Life?
Never
Less often
2-3 times a month
Once a week
Several times a week
Once a day
Several times a day
How many Linden Dollars do you earn in Second Life per week?
L$0
L$1-100
L$101-500
L$501-1,000
More than L$1,000
No information
How many Linden Dollars do you spend in Second Life per week?
L$0
L$1-100
L$101-500
L$501-1,000
More than L$1,000
No information
18
40
73
92
235
122
3.1%
6.9%
12.6%
15.9%
40.5%
21.0%
50
114
95
124
135
62
8.6%
19.7%
16.4%
21.4%
23.3%
10.7%
23
64
92
119
173
55
54
4.0%
11.0%
15.9%
20.5%
29.8%
9.5%
9.3%
31
117
143
121
107
37
24
5.3%
20.2%
24.7%
20.9%
18.4%
6.4%
4.1%
99
108
169
74
94
36
17.1%
18.6%
29.1%
12.8%
16.2%
6.2%
58
134
177
86
90
35
10.0%
23.1%
30.5%
14.8%
15.5%
6.0%
Flagship Brand Stores within Virtual Worlds
moderation” in Tables 4-7). In case moderation was
indicated on an overall level, we subsequently identified specific moderating effects by constraining individual paths to be equal across groups (while letting all
other paths vary freely) and compared the change in
model chi-square to a fully unconstrained model (See
panel “test of individual paths” in Tables 4-7). In a
last step, we determined path coefficient estimates
and standard errors for a model in which we constrained all paths to be equal across groups unless a significant moderating effect had been observed.
RESULTS
Table 3 shows the estimation results (parameter
estimates, standard errors, and associated p-values)
as well as goodness-of-fit indices for the baseline
model estimated using all 580 observations. With
respect to model fit, the Comparative Fit Index (CFI)
as well as the Tucker-Lewis Index (TLI) exceed 0.95
and the Root Mean Squared Error of Approximation
(RMSEA) and Standardized Root Mean Squared
Residual (SRMR) are both below the recommended
thresholds of 0.06 and 0.08, respectively (Hu and
Bentler, 1999). Combined, this indicates excellent
model fit. Out of the six paths estimated, four are
significant at the 5% level. Attitude toward the shopkeeper and mood positively and significantly
influence a user’s attitude toward a Second Life flagship store (parameter estimates of 0.651 and 0.159,
respectively). Attitude toward a Second Life flagship
store subsequently positively influences the attitude
toward the real-life brand (parameter estimate:
0.392), which in turn positively impacts real-life purchase intent (parameter estimate: 0.571). Combined,
this leads to confirmation of H1b, H1d, H2 and H3.
There is, however, no significant influence of store
credibility and attitude toward purchasing on Second
Life on store attitudes. H1a and H1c are therefore
rejected on an overall level. This result implies that
exposure to a Second Life flagship store influences
real-life brand attitudes and purchase intent and therefore provides an indication of the existence of spill-
67
over effects between Second Life and real life,
consistent with our conceptual framework.
To test the robustness of these findings with respect
to alternative model specifications, we also estimated a
model that did not include mood as an antecedent of
store attitudes. It might be conceivable that our
respondents were not able to accurately recall the
mood they were in when visiting the respective
Second Life flagship store and that this lack of accuracy resulted in biased estimates.6 Although such a
reduced model fits the data slightly worse than our
baseline model (CFI: 0.977, TLI: 0.970, RMSEA:
0.050, SRMR: 0.060), our main implications remain
unchanged: Store credibility and attitude toward purchasing on Second Life do not show a significant
impact on store attitudes, while attitude toward the
shopkeeper influences store attitudes positively and
significantly. More importantly, also in this model we
observe a significant and positive relationship between store attitudes and brand attitudes (parameter
estimate: 0.395) and between brand attitudes and
real-life purchase intent (parameter estimate: 0.576).
We therefore conclude that our findings remain
robust to the exclusion of mood as a store attitude
antecedent.
To test for a potential moderating impact of usage
intensity, purchase experience and use gratification
on these relationships, we proceeded as outlined
above and first performed an omnibus test of moderation for each construct. For usage intensity, we analyzed two variables, usage duration and weekly usage,
each measured by one question (“When did you start
using Second Life?” and “How many hours do you
spend per week on Second Life?”, respectively).7
The difference in model chi-square between a
constraint and free model is 16.397 for usage duration and 16.744 for weekly usage, which turns out to
be insignificant on 12 df (p-values of 0.1737 and
0.1595, respectively). Combined, this indicates that
usage intensity does not exert a moderating impact
on the relationships visualized in Figure 1, leading to
rejection of H4a and H4b.
6. We thank an anonymous reviewer for bringing this point to our
attention.
7. The distribution of respondents along these two variables were as
follows: For usage duration: less than 6 months ago – 223 respondents; 6-12 months ago – 235 respondents; more than 12 months ago
– 122 respondents. For weekly usage: less than 9 hours – 164
respondents; 10-24 hours – 219 respondents; more than 24 hours –
197 respondents.
Att. (SLFS)
Att. (SLFS)
Att. (SLFS)
Att. (SLFS)
Att. (RL Brand)
RL PI
Att. (SLFS)
Att. (SLFS)
Att. (SLFS)
Att. (SLFS)
Att. (RL Brand)
RL PI
→
→
→
→
→
→
→
→
→
→
→
→
Att. (Shopkeeper)
Att. (SL purchasing)
Store credibility
Mood
Att. (SLFS)
Att. (RL Brand)
Att. (Shopkeeper)
Att. (SL purchasing)
Store credibility
Mood
Att. (SLFS)
Att. (RL Brand)
Model fit indices
542.554
Model Chi2
Model df
238
CFI
0.976
TLI
0.970
RMSEA
0.047
SRMR
0.055
Baseline model
Standard Error
0.059
0.041
0.048
0.036
0.055
0.051
p-value
0.0000
0.0750
0.6617
0.0000
0.0000
0.0000
Overview of moderating effects (p-value)
Shopping
Purchase
Spent per
frequency
frequency
purchase
Use gratification
0.0585
0.0001
0.2527
0.0010
0.5758
0.0083
0.1505
0.2080
0.2169
0.0127
0.1476
0.0080
0.0381
0.0956
0.0994
0.0005
0.0547
0.2257
0.0011
0.2153
0.4500
0.4251
0.9738
0.6275
Estimate
0.651
– 0.073
0.021
0.159
0.392
0.571
Table 3. – Model estimation results – Baseline model
68
Michael Haenlein, Andreas M. Kaplan
Flagship Brand Stores within Virtual Worlds
For purchase experience, we analyzed three
variables, shopping frequency, purchase frequency,
and spending per purchase, each measured by one
question (“How often do you go shopping/make a
purchase on Second Life?” and “How many Linden
Dollars do you spend on average when you make a
purchase on Second Life?”). As can be seen in Tables
4, 5, and 6, the omnibus test provides a significant
indication of moderation effect in all three cases
(p-values above or equal to 0.0317). Looking at the
specific paths moderated by these variables, it can be
seen that shopping frequency moderates the impact
of mood on store attitude (p-value: 0.0381), purchase
frequency the impact of attitude toward the shopkeeper, attitude toward purchasing on Second Life, and
store credibility on store attitude (p-values above or
equal to 0.0127), and spending per purchase the relationship between store and brand attitude (p-value:
0.0011). Combined, these findings support H5a and
H5b. For use gratification, Table 7 shows that the
omnibus test indicates significant moderation (pvalue: 0.0002) and the analysis of individual paths
reveals that the impact of attitude toward the shopkeeper, store credibility, and mood on store attitudes
differ significantly across use gratification clusters
(p-values above or equal to 0.0080). This confirms
H6a and leads to rejection of H6b.
To better understand the nature of these moderating
effects, we determined group-specific parameter estimates for each path where significant moderation
had been detected. The results of this analysis can be
found in Tables 4-7. For ease of interpretation, we
present our findings by type of relationship moderated
rather than type of moderating variable. The impact
of attitude toward the shopkeeper on store attitude
follows an inverse U-shaped relationship for purchase frequency. While path coefficients are relatively low for low and high levels of purchase frequency (0.291 and 0.264, respectively), they increase
for medium levels (0.549). An opposite pattern
emerges for use gratification, where the relationship
between attitude toward the shopkeeper and store
attitudes is stronger for light and heavy users (0.741
and 1.009, respectively) than for medium ones
(0.524). Attitude toward purchasing on Second Life,
which was insignificant on the overall level, turns out
to significantly influence store attitudes for high
levels of purchase frequency. However, different to
what might be expected, the relationship between the
69
two variables is a negative one (path coefficient:
–0.151), implying that more favorable attitudes
toward purchasing on Second Life lead to less favorable store attitudes. Store credibility, which was also
insignificant in the baseline model, has a significant
and positive relationship to store attitudes for high
levels of purchase frequency (0.116) and for medium
levels of use gratification (0.248). Mood exerts a
significant influence on store attitudes only for
medium levels of shopping frequency (0.220) and
light/medium use gratification clusters (0.208 and
0.253). For all other groups (low and high levels of
shopping frequency as well as heavy use gratification
users), the relationship, which received significant
support in the baseline model, becomes insignificant.
Finally, the relationship between store attitudes and
brand attitudes becomes insignificant for medium
levels of purchase spending (p-value 0.1323), but
remains significant and positive for low and high
levels (path coefficients of 0.622 and 0.322, respectively).
DISCUSSION AND MANAGERIAL IMPLICATIONS
To sum up, our analysis results in the following
three findings: First, our study provides a contribution to developing a better understanding of the new
consumer. Over the past decade, the widespread use of
the Internet has resulted in a new range of contact
channels and it has been stated that new models
might be needed to explain consumer behavior
within such an environment. Our study provides two
types of insight with respect to this question: First,
we confirm a statement that has previously been
made in a different setting, namely that consumers do
not separate their online from their offline activities
but instead consider both as forms of self-expression.
Consistent with consumer culture theory (Arnould
and Thompson, 2005), consumers use the Internet to
develop virtual self-presentations (Schau and Gilly,
2003). Experiences made online are subsequently
transferred offline and vice versa, as can be seen,
among others, in our study. From a company pers-
Never
Less often
2-3 times a month
Group B Once a week
Several times a week
Once a day
Group C
Several times a day
Chi2
1,575.479
1,570.904
1,572.857
1,576.333
1,575.611
1,571.397
Att. (Shopkeeper) → Att. (SLFS)
Att. (SL purchasing) → Att. (SLFS)
Store credibility
→ Att. (SLFS)
Mood
→ Att. (SLFS)
Att. (SLFS)
→ Att. (RL Brand)
Att. (RL Brand)
→ RL PI
**: p < 0.01 *: p < 0.05
50%
19%
292
109
580 100%
31%
in %
179
N
878
878
878
878
878
878
df
5.679
1.104
3.057
6.533
5.811
1.597
2
2
2
2
2
2
0.0585
0.5758
0.2169
0.0381
0.0547
0.4500
∆ Chi2 ∆ df p-value
df ∆ Chi2 ∆ df p-value
Chi2
1,592.358 888
1,569.800 876 22.558 12 0.0317
Test of individual paths
Omnibus test of moderation
Constrained (all paths)
Free
Group A
N
23
64
92
119
173
55
54
580
Shopping frequency:
How often do you go shopping in Second Life?
Group-specific parameter estimates
Group A
Group B
Group C
(Low)
(Medium)
(High)
0.660**
0.660**
0.660**
– 0.068
– 0.068
– 0.068
0.024
0.024
0.024
0.101
0.220**
0.122
0.391**
0.391**
0.391**
0.572**
0.572**
0.572**
Table 4. – Test of moderating effects – Shopping frequency
70
Michael Haenlein, Andreas M. Kaplan
Chi2
1,639.673
1,630.930
1,630.092
1,626.054
1,624.335
1,623.069
Att. (Shopkeeper) → Att. (SLFS)
Att. (SL purchasing) → Att. (SLFS)
Store credibility
→ Att. (SLFS)
Mood
→ Att. (SLFS)
Att. (SLFS)
→ Att. (RL Brand)
Att. (RL Brand)
→ RL PI
**: p < 0.01 *: p < 0.05
46%
29%
264
168
580 100%
26%
in %
148
N
2
2
2
2
2
2
0.0001
0.0083
0.0127
0.0956
0.2257
0.4251
∆ Chi2 ∆ df p-value
878 18.315
878 9.572
878 8.734
878 4.696
878 2.977
878 1.711
df
df ∆ Chi2 ∆ df p-value
Chi2
1,654.454 888
1,621.358 876 33.096 12 0.0009
N
31
117
143
121
107
37
24
580
Test of individual paths
Omnibus test of moderation
Constrained (all paths)
Free
Never
Group A
Less often
2-3 times a month
Group B Once a week
Several times a week
Group C Once a day
Several times a day
Purchase frequency:
How often do you make a purchase in Second Life?
Group-specific parameter estimates
Group A
Group B
Group C
(Low)
(Medium)
(High)
0.291**
0.549**
0.264**
– 0.024
0.006
– 0.151**
0.024
– 0.068
0.016*
0.072**
0.072**
0.072**
0.374**
0.374**
0.374**
0.581**
0.581**
0.581**
Table 5. – Test of moderating effects – Purchase frequency
Flagship Brand Stores within Virtual Worlds
71
$L0-25
$L25-100
Group B $L100-300
$L300-500
Group C
More than $L500
Chi2
1,515.583
1,516.619
1,516.659
1,517.450
1,526.446
1,512.885
Att. (Shopkeeper) → Att. (SLFS)
Att. (SL purchasing) → Att. (SLFS)
Store credibility
→ Att. (SLFS)
Mood
→ Att. (SLFS)
Att. (SLFS)
→ Att. (RL Brand)
Att. (RL Brand)
→ RL PI
**: p < 0.01 *: p < 0.05
30%
32%
176
183
580 100%
38%
in %
221
N
2
2
2
2
2
2
0.2527
0.1505
0.1476
0.0994
0.0011
0.9738
∆ Chi2 ∆ df p-value
878 2.751
878 3.787
878 3.827
878 4.618
878 13.614
878 0.053
df
df ∆ Chi2 ∆ df p-value
Chi2
1,538.483 888
1,512.832 876 25.651 12 0.0120
Test of individual paths
Omnibus test of moderation
Constrained (all paths)
Free
Group A
N
104
117
176
104
79
580
Group-specific parameter estimates
Group A
Group B
Group C
(Low)
(Medium)
(High)
0.639**
0.639**
0.639**
– 0.074
– 0.074
– 0.074
0.017
0.017
0.017
0.155**
0.155**
0.155**
0.622**
0.149
0.322**
0.572**
0.572**
0.572**
Spent per purchase:
How many Linden Dollars do you spend on average when you make a purchase on Second Life?
Table 6. – Test of moderating effects – Spending per purchase
72
Michael Haenlein, Andreas M. Kaplan
Group A
225
39%
Group A
– 0.34
0.19
– 0.61
– 1.24
Chi2
1,753.668
1,743.051
1,749.575
1,755.317
1,742.982
1,740.843
Att. (Shopkeeper) → Att. (SLFS)
Att. (SL purchasing)→ Att. (SLFS)
Store credibility → Att. (SLFS)
Mood
→ Att. (SLFS)
Att. (SLFS)
→ Att. (RL Brand)
Att. (RL Brand)
→ RL PI
**: p < 0.01 *: p < 0.05
Group C
106
18%
Group C
1.81
2.23
1.31
– 0.93
580
100%
Total
0.64
1.09
0.33
– 1.00
2
2
2
2
2
2
0.0010
0.2080
0.0080
0.0005
0.2153
0.6275
∆ Chi2 ∆ df p-value
878 13.757
878 3.140
878 9.664
878 15.406
878 3.071
878 0.932
df
df ∆ Chi2 ∆ df p-value
Chi2
1,776.750 888
1,739.911 876 36.839 12 0.0002
Group B
249
43%
Group B
1.03
1.42
0.75
– 0.82
Test of individual paths
Omnibus test of moderation
Constrained (all paths)
Free
Diversion
Personal relationships
Personal identity
Earning money
N
in %
Use gratification:
Group-specific parameter estimates
Group A
Group B
Group C
(Low)
(Medium)
(High)
0.741**
0.524**
1.009**
– 0.097
– 0.097
– 0.097
– 0.011
0.248**
– 0.093
0.208**
0.253**
– 0.154
0.381**
0.381**
0.381**
0.571**
0.571**
0.571**
Table 7. – Test of moderating effects – Use gratification
Flagship Brand Stores within Virtual Worlds
73
74
Michael Haenlein, Andreas M. Kaplan
pective, this tight integration in the mind of the
consumer implies that traditional communication
strategies need to be closely aligned with activities
within virtual worlds or other online channels to
avoid conflicting messages and inconsistencies. This is
especially important as the time that users spend
within Second Life can be substantial (i.e., more than
10 hours per week for 70% of our respondents and
more than 24 hours per week for 33%). Second, we
provide an indication that traditional models of
consumer behavior prove highly robust when being
applied to a new setting. The AAd – ABrand relationship
which underlies our conceptual framework has first
been proposed more than 25 years ago (Shimp, 1981)
and proves to be accurate even in the setting of virtual
worlds. The four different store attitude antecedents,
which we derived based on literature in traditional
advertising research, explain 62.2% of the variance
in the store attitude variable. Combined, this leads to
the assumption that consumer behavior within the
“new” environment consumers are facing today may
not be fundamentally different from behavior observed
in traditional settings and that well-established
models might still turn out to be highly useful, even
when applied outside their traditional scope. While
this finding may be surprising, and even unintuitive, as
it indicates that users behave the same way in virtual
and real worlds, it can be seen as an indication of the
strong future business potential of virtual environments for conducting marketing research and virtual
commerce.
Second, we show that there is a significant and
positive relationship between a user’s attitude toward a
Second Life flagship store and the user’s attitude
toward the real-life brand. This implies that virtual
flagship brand stores can be expected to fulfill functions similar to (online) advertising, besides their
potential role as distribution hubs for virtual products
and services. Our work therefore confirms the common belief that activities within virtual worlds such
as Second Life can be used to support an overall
branding strategy (Enright, 2007). This finding is of
managerial relevance for two reasons: First, many
firms entering Second Life have realized that the
revenue that can be achieved by selling virtual products and services does not compensate for the cost of
setting up a virtual flagship store. A private region
within Second Life of 65,536 m2 is, for example,
sold for US$1,000 plus US$295 monthly mainte-
nance fee and the cost of designing a flagship store
can range from several hundred dollars up to
U$200,000 (Kaplan and Haenlein, 2009c). This
investment is difficult to recover through the sales of
digital products alone as the willingness-to-pay of
Second Life residents remains limited (e.g., approximately L$300 or US$1.20 for a digital suit). Our analysis provides an indication that entering Second Life
may nevertheless be a profitable strategy when
taking the potential advertising and spill-over effects to
real life into account. Second, our study is one of the
first that empirically investigates the efficacy of corporate activities within the larger group of social
media (Kaplan and Haenlein, 2010). Social media
are a domain of increasing importance for many
firms nowadays but not much is known about their
efficacy and return-on-investment. Our analysis can
be seen as the first step to develop a better understanding of these new contact channels.
Finally, we provide an indication that the advertising impact of Second Life flagship stores is not the
same for all users but instead varies with increasing
levels of purchase frequency or use gratification
sought from Second Life usage. Some of the effects
we observe, such as the significant impact of store
credibility or attitude toward purchasing on Second
Life on store attitude at high levels of purchase frequency, are consistent with a notion of learning in
which meta-knowledge is created only after sufficient experience. Other results, such as the (inverse)
U-shaped relationship of attitude toward the shopkeeper on store attitude with increasing levels of purchase frequency/use gratification or the varying
impact of mood can be explained with the assumption that Second Life residents evolve during their
Second Life usage and therefore allocate different
decision weights to the same variables depending on
their degree of exposure to the medium. Potentially,
such an evolution takes place because for experienced users Second Life is not considered as a mere
computer game anymore but as an extension of the
user’s real life, in line with the results obtained by
Kaplan and Haenlein (2009a). However, the lack of
longitudinal design within our data collection does
not allow the testing of such a hypothesis in more
detail. From a managerial perspective, this implies
that companies may need to target Second Life residents differently, depending on their purchase frequency/use gratification. Hence, Second Life flag-
Flagship Brand Stores within Virtual Worlds
ship stores that are located in areas with a disproportionate share of experienced residents (e.g., residential
areas in which most residents own real estate) may
need to be designed differently from presences in
areas that mainly attract new users (e.g., shopping
malls) in order to be effective. Also, flagship stores
whose primary intention is to serve as advertising
media should be targeted toward a different group of
Second Life residents than those who are mainly
designed to sell virtual products and services. The
fact that the relationship between store attitude and
brand attitude becomes insignificant for medium
levels of purchase spending shows that the advertising effectiveness of Second Life may disappear for
some user groups.
LIMITATIONS AND AREAS OF FUTURE RESEARCH
Obviously, our analysis can only be seen as a first
step to better understand the consumer use and business potential of virtual social worlds in general and
Second Life in particular. Future research could, for
example, address some of the limitations in our
approach and extend our study by an analysis of the
impact of Second Life flagship stores on actual
consumer behavior (vs. behavioral intentions) or a
longitudinal investigation of the impact of multiple
exposures to the same flagship stores. Also, the fact
that the survey participants in our study were able to
choose the Second Life flagship store to visit themselves may have introduced a bias in our analysis as
residents who are more or less involved with Second
Life may use different choice criteria to decide which
store to visit. Addressing these points would strengthen
our key finding that exposure to Second Life flagship
stores can spill over into real life and provide further
insight into the advertising efficacy of this new
medium.
Another interesting question that merits future
research is whether the creation of Second Life flagship stores, and especially the distribution of virtual
equivalents of real life products within Second Life,
could also have negative consequences in terms of
75
real-life purchase intentions and sales. If companies
use their flagship stores to distribute virtual products
that are similar to their real-life offerings, but differ in
certain features, it might be possible that users form
unrealistic expectations for real-life products based
on their Second Life experience. Technically, it is for
example not difficult to offer fashion items in a very
large variety of colors within Second Life (much larger
than would be feasible in real life) or with functionalities that could not exist in real life (e.g., flying shoes or
invisibility cloaks). However, by doing so, companies could artificially raise expectations with respect to
the real-life products offered by the company that are
impossible to fulfill. In the context of the expectancy
confirmation framework (Oliver, 1980), such unfulfilled expectations could lead to feelings of dissatisfaction, deception, and ultimately lower real-life purchase intention and sales. On the other hand, simply
replicating a company’s real-life offering within
Second Life might be perceived as uninventive and
boring by Second Life residents, potentially leading to
negative (real-life) brand associations. Current managerial belief appears to be that entering Second Life is
a low-risk strategy as doing so either has a positive or
neutral impact on real-life sales. However, a better
understanding of the likely risks associated with such a
strategy is needed in order to determine the optimal
use of Second Life in a given situation.
Finally, future studies could also take the company (vs. consumer) perspective and analyze how
firms in different industries currently use virtual
social worlds. In a first step, it might, for example, be
interesting to investigate for what types of activities
companies are relying on these new media, and what
objectives they plan to achieve with their usage. In a
next step, it would then be worthwhile to investigate
how the market reacts to the decision to open a corporate presence in any of these virtual social worlds. In a
similar context, Geyskens, Gielens and Dekimpe
(2002) applied an event-study methodology to investigate the stock market reaction to the introduction of eCommerce sites. Studies in that spirit could be
undertaken to determine the ultimate value creation
potential of performing corporate activities within
Second Life.
76
Michael Haenlein, Andreas M. Kaplan
BIBLIOGRAPHICAL REFERENCES
Arnould E.J. and Thompson C.J. (2005), Consumer
Culture Theory (CCT): Twenty years of research,
Journal of Consumer Research, 31, 4, 868-882.
Baker M.J. and Churchill G.A.J. (1977), The impact of
physically attractive models on advertising evaluations,
Journal of Marketing Research, 14, 4, 538-555.
Barnes S. and Mattsson J. (2008), Brand value in virtual
worlds: an axiological approach, Journal of Electronic
Commerce Research, 9, 3, 195-206.
Birnbaum M.H. (2004), Human research and data collection via the Internet, Annual Review of Psychology, 55, 1,
803-832.
Brakus J.J., Schmitt B.H. and Zarantonello L. (2009), Brand
experience: What is it? How is it measured? Does it
affect loyalty?, Journal of Marketing, 73, 3, 52-68.
Chung D.S. and Kim S. (2008), Blogging activity among
cancer patients and their companions: uses, gratifications, and predictors of outcomes, Journal of the
American Society for Information Science and
Technology, 59, 2, 297-306.
Cox E.P.I. (1980), The optimal number of response alternatives for a scale: a review, Journal of Marketing
Research, 17, 4, 407-422.
Doyle S.A., Moore C.M., Doherty A.M. and Hamilton M.
(2008), Brand context and control: the role of the flagship store in B&B Italia, International Journal of
Retail & Distribution Management, 36, 7, 551-563.
Edvardsson B., Enquist B. and Johnston R. (2005),
Cocreating customer value through hyperreality in the
prepurchase service experience, Journal of Service
Research, 8, 2, 149-161.
Enright A. (2007), How the second half lives, Marketing
News, 41, 3, 12-14.
Erdem T. and Swait J. (2004), Brand credibility, brand
consideration and choice, Journal of Consumer
Research, 31, 1, 191-198.
Firat A.F., Sherry J.F.J. and Venkatesh A. (1994),
Postmodernism, marketing and the consumer,
International Journal of Research in Marketing, 11, 4,
311-316.
Firat A.F. and Venkatesh A. (1993), Postmodernity: The
age of marketing, International Journal of Research in
Marketing, 10, 3, 227-249.
Gardner M.P. (1985a), Does attitude toward the ad affect
brand attitude under a brand evaluation set?, Journal of
Marketing Research, 22, 2, 192-198.
Gardner M.P. (1985b), Mood states and consumer behavior: a critical review, Journal of Consumer Research,
12, 3, 281-300.
Gefen D., Karahanna E. and Straub D.W. (2003), Trust and
TAM in online shopping: an integrated model, MIS
Quarterly, 27, 1, 51-90.
Geyskens I., Gielens K. and Dekimpe M.G. (2002), The
market valuation of Internet channel additions, Journal
of Marketing, 66, 2, 102-119.
Grayson K. and Martinec R. (2004), Consumer perceptions
of iconicity and indexicality and their influence on
assessments of authentic market offerings, Journal of
Consumer Research, 31, 2, 296-312.
Ha L. (2008), Online advertising research in advertising
journals: a review, Journal of Current Issues and
Research in Advertising, 30, 1, 31-48.
Hovland C.I. and Weiss W. (1951), The influence of source
credibility on communication effectiveness, Public
Opinion Quarterly, 15, 4, 635-650.
Hu L.-T. and Bentler P.M. (1999), Cutoff criteria for fit
indexes in covariance structure analysis: conventional
criteria versus new alternatives, Structural Equation
Modeling, 6, 1, 1-55.
Iyer R. and Eastman J.K. (2006), The elderly and their attitudes toward the Internet: the impact on Internet use,
purchase, and comparison shopping, Journal of
Marketing Theory and Practice, 14, 1, 57-67.
Kang H., Hahn M., Fortin D.R., Hyun Y.J. and Eom Y.
(2006), Effects of perceived behavioral control on the
consumer usage intention of e-coupons, Psychology &
Marketing, 23, 10, 841-864.
Kaplan A.M. (2009), Second Life: leçons pour le monde
réel, Expansion Management Review, 133, June, 58-60.
Kaplan A.M. and Haenlein M. (2009a), Consumer use and
business potential of virtual worlds: the case of Second
Life, The International Journal on Media Management,
11, 3.
Kaplan A.M. and Haenlein M. (2009b), Consumers, companies and virtual social worlds: a qualitative analysis of
Second Life, Advances in Consumer Research, 36, 1,
873-874.
Kaplan A.M. and Haenlein M. (2009c), The fairyland of
Second Life: about virtual social worlds and how to
use them, Business Horizons, 52, 6.
Kaplan A.M. and Haenlein M. (2009d), Utilisation et
potentiel commercial des hyperréalités : une analyse
qualitative de Second Life, Revue Française du
Marketing, 222, 69-81.
Kaplan A.M. and Haenlein M. (2010), Users of the world,
unite! The challenges and opportunities of social
media, Business Horizons, forthcoming.
Katz E., Blumler J.G. and Gurevitch M. (1973-1974), Uses
and gratifications research, Public Opinion Quarterly,
37, 4, 509-523.
Ko H., Cho C.-H. and Roberts M.S. (2005), Internet uses
and gratifications: a structural equation model of interactive advertising, Journal of Advertising, 34, 2, 57-70.
Kozinets R.V., Sherry J.F., DeBerry-Spence B., Duhachek
A., Nuttavuthisit K. and Storm D. (2002), Themed
flagship brand stores in the new millennium: theory,
practice, prospects, Journal of Retailing, 78, 1, 17-29.
Lemmens J.S. and Bushman B.J. (2006), The appeal of
violent video games to lower educated aggressive adolescent boys from two countries, CyberPsychology &
Behavior, 9, 5, 638-641.
Liang T.-P., Lai H.-J. and Ku Y.-C. (2006-2007),
Personalized content recommendation and user satisfaction: theoretical synthesis and empirical findings,
Flagship Brand Stores within Virtual Worlds
Journal of Management Information Systems, 23, 3,
45-70.
Lin A.C. (2008), Virtual consumption: a Second Life for
earth?, Brigham Young University Law Review, 1, 47114.
MacKenzie S.B. and Lutz R.J. (1989), An empirical examination of the structural antecedents of attitude toward
the ad in an advertising pretesting context, Journal of
Marketing, 53, 2, 48-65.
MacKenzie S.B. and Spreng R.A. (1992), How does motivation moderate the impact of central and peripheral
processing on brand attitudes and intentions?, Journal of
Consumer Research, 18, 4, 519-529.
Manchanda P., Dube J.-P., Goh K.Y. and Chintagunta P.K.
(2006), The effect of banner advertising on Internet
purchasing, Journal of Marketing Research, 43, 1, 98108.
McQuail D., Blumler J.G. and Brown J.R. (1972), The
television audience: a revised perspective, in D.
McQuail (Ed.), Sociology of mass communication,
Harmondsworth, Penguin Books.
McSweeney F.K. and Bierley C. (1984), Recent developments in classical conditioning, Journal of Consumer
Research, 11, 2, 619-631.
Mitchell A.A. and Olson J.C. (1981), Are product attribute
beliefs the only mediator of advertising effects on
brand attitude?, Journal of Marketing Research, 18, 3,
318-332.
Muthén B.O. (2002), Beyond SEM: general latent variable
modeling, Behaviormetrika, 29, 1, 81-117.
Muthén L.K. and Muthén B.O. (1998-2007), Mplus User’s
Guide (5th Edition). Los Angeles, CA, Muthén &
Muthén.
Oliver R.L. (1980), A cognitive model of the antecedents
and consequences of satisfaction decisions, Journal of
Marketing Research, 17, 4, 460-469.
Podsakoff P.M., MacKenzie S.B., Lee J.-Y. and Podsakoff
N.P. (2003), Common method bias in behavioral
research: a critical review of the literature and recommended remedies, Journal of Applied Psychology, 88, 5,
879-903.
Qiu G. and Papatla P. (2008), An empirical analysis of
inter-acquisition time of free online content, Journal of
Interactive Marketing, 22, 2, 19-27.
Rose R.L. and Wood S.L. (2005), Paradox and the
consumption of authenticity through reality television,
Journal of Consumer Research, 32, 2, 284-296.
Schau H.J. and Gilly M.C. (2003), We are what we post?
Self-presentation in personal web space, Journal of
Consumer Research, 30, 3, 385-404.
77
Schillewaert N. and Meulemeester P. (2005), Comparing
response distributions of offline and online data collection methods, International Journal of Market
Research, 47, 2, 163-178.
Schlosser A.E. (2003), Experiencing products in the virtual
world: the role of goal and imagery in influencing attitudes versus purchase intentions, Journal of Consumer
Research, 30, 2, 184-198.
Schlosser A.E. (2006), Learning through virtual product
experience: the role of imagery on true versus false
memories, Journal of Consumer Research, 33, 3,
377-383.
Shimp T.A. (1981), Attitude toward the ad as a mediator of
consumer brand choice, Journal of Advertising, 10, 2,
9-15.
Spence J. (2008), Editor’s introduction, Journal of Virtual
Worlds Research, 1, 1, 1.
Stafford T.F., Stafford M.R. and Schkade L.L. (2004),
Determining uses and gratifications for the Internet,
Decision Sciences, 35, 2, 259-288.
Stuart E.W., Shimp T.A. and Engle R.W. (1987), Classical
conditioning of consumer attitudes: four experiments
in an advertising context, Journal of Consumer
Research, 14, 3, 334-349.
Swinyard W.R. (1993), The effects of mood, involvement,
and quality of store experience on shopping intentions,
Journal of Consumer Research, 20, 2, 271-280.
Vargo S.L. and Lusch R.F. (2004), Evolving to a new
dominant logic for Marketing, Journal of Marketing,
68, 1, 1-17.
Venkatesh A., Sherry J.F.J. and Firat A.F. (1993),
Postmodernism and the marketing imaginary,
International Journal of Research in Marketing, 10, 3,
215-223.
Wofford J.C. and Goodwin V.L. (1990), Effects of feedback on cognitive processing and choice of decision
style, Journal of Applied Psychology, 75, 6, 603-612.
Wood N.T., Solomon M.R. and Allan D. (2008), Welcome to
the matrix: e-learning gets a Second Life, Marketing
Education Review, 18, 2, 47-53.
Yee N. and Bailenson J. (2007), The Proteus effect: the
effect of transformed self-representation on behavior,
Human Communication Research, 33, 3, 271-290.
Yee N., Bailenson J.N., Urbanek M., Chang F. and Merget
D. (2007), The unbearable likeness of being digital: the
persistence of nonverbal social norms in online virtual
environments, CyberPsychology & Behavior, 10, 1,
115-121.
78
Michael Haenlein, Andreas M. Kaplan
Appendix. – Measurement scales
All measures employ 7-point scales with “strongly disagree/agree” as anchors, except where noted otherwise.
Store credibility (Erdem and Swait, 2004, Cronbach’s Alpha: 0.856)
• The SL presence of BRAND reminds me of someone who is competent and knows what s/he is doing
• The SL flagship store of BRAND delivers what it promises
• The claims made by the SL flagship store of BRAND are believable
• Over time, my experiences with the SL flagship store of BRAND have led me to expect it to keep its promises,
no more and no less
• The SL flagship store of BRAND doesn’t pretend to be something it isn’t
Attitude toward purchasing on Second Life (Iyer and Eastman, 2006, Cronbach’s Alpha: 0.671)
• Buying on SL is more fun than traditional buying
• I enjoy buying on SL
• Buying on SL is no riskier than traditional buying
• I am confident in my ability to buy successfully on SL
Attitude toward the shopkeeper (MacKenzie and Lutz, 1989, Cronbach’s Alpha: 0.951)
• How would you evaluate the company that owns the SL flagship store of BRAND?
– Bad/good
– Unpleasant/pleasant
– Unfavorable/favorable
Mood (Swinyard, 1993, Cronbach’s Alpha: 0.929)
• Indicate to what extent you have felt this way while you were visiting the SL flagship store of BRAND
– Sad/happy
– Bad mood/good mood
– Irritable/pleased
– Depressed/cheerful
Attitude toward the Second Life flagship store (MacKenzie and Lutz, 1989, Cronbach’s Alpha: 0.963)
• Below you will find three pairs of adjectives. Indicate how well one or other adjective in each pair describes
how you perceived the SL flagship store of BRAND
– Bad/good
– Unpleasant/pleasant
– Unfavorable/favorable
Attitude toward the Real Life brand (Gardner 1985a, Cronbach’s Alpha: 0.963)
• Below you will find three pairs of adjectives. Indicate how well one or other adjective in each pair describes
your overall feeling toward BRAND
– Bad/good
– Unpleasant/pleasant
– Dislike/like
Real Life purchase intent (Baker and Churchill, 1977, Cronbach’s Alpha: 0.875)
• How likely would you be to try BRAND in Real Life?
• How likely would you be to buy BRAND if you happened to see it in a real-life store?
• How likely would you be to actively seek out BRAND in a real-life store in order to purchase it?
Flagship Brand Stores within Virtual Worlds
Use gratification (McQuail et al., 1972)
• Diversion (Cronbach’s Alpha: 0.710)
– Being in Second Life helps me to escape from the boredom of everyday live
– Being in Second Life takes me out of myself
– Being in Second Life helps me to get away from my problems
• Personal relationships (Cronbach’s Alpha: 0.694)
– The people in Second Life have become like close friends to me
– Being in Second Life is good company when you’re alone
– After having been in Second Life, I look forward to talking about it with others
– After having been in Second Life, it is a topic of conversation
• Personal identity (Cronbach’s Alpha: 0.725)
– Being in Second Life reminds me of things that have happened in my own life
– Being in Second Life sometimes brings back memories of certain people I used to know
– Being in Second Life sometimes helps me to understand my own life
– Being in Second Life provides food for thought
• Being in Second Life helps me to earn Real Life money
79
Download